Our objective is to further characterize the natural history of clinically localized prostate cancer managed conservatively. From standard prognostic factors, we will construct prognostic nomograms to compute for individual patients the probabilities of clinical progression (symptomatic local or distant) and of cancer-related death at a given time after diagnosis. Information from these "conservative management" nomograms and our prior aggressive therapy nomograms will be used to construct more robust decision analysis models for men with clinically localized prostate cancer. Through a prospective clinical trial, we will test the hypothesis that men who choose treatment recommended by this model will be less likely to regret their decision. We will also use novel gene expression methods to identify and validate prognostically useful molecular markers that have the potential to increase the predictive accuracy of our nomograms. Our Specific Aims are: 1) To further define the natural history of clinically localized prostate cancer by identifying in an epidemiologic, population-based study in the United Kingdom a consecutive cohort of 2000 men diagnosed with prostate cancer in the PSA era but managed conservatively. In collaboration with the Imperial Cancer Research Fund, we will identify from regional tumor registries 2000 men diagnosed between 1990 and 1996 who had no treatment for at least 6 months after diagnosis and are eligible for a minimum follow-up of 5 years (maximum 10-16 years by completion of the SPORE in 2006). We will identify a second cohort of 2000 men treated with hormonal therapy alone within 6 months of diagnosis. Diagnostic biopsy specimens (mostly TURP) and clinical information will be retrieved and uniformly graded and staged with centralized review and the data computerized for further analyses. 2) To derive a nomogram that predicts the probability of progression (to symptomatic local or distant metastases), based on clinical prognostic factors (clinical stage, Gleason grade, and PSA level) for those cohorts of men managed conservatively (untreated or with hormonal therapy alone). 3) To construct a decision-analysis model which incorporates information from the conservative management nomograms and the aggressive-therapy nomograms previously developed and determine in a prospective clinical trial whether patients who chose treatment identified by the model as most beneficial (in quality adjusted life years) are less likely to regret their choice. 4) To identify and validate expressed genes in prostate cancer that correlate with outcome and evaluate their ability to complement nomogram prediction of progression for clinically localized disease.